The use of decomposition SVD to approximate a surface

Abstract

This paper describes the procedure of determining the parameters of the approximating function of the surface using a distribution of special value. From a practical point of view, an important issue is to determine the covariance matrix of the estimated parameters. Interval estimation was carried out and a methodology to obtain an optimal equation of approximating surface was presented. The main emphasis was given to the elimination of these parameters functions that generate unnecessary disturbance. The results obtained using the approximate decomposition SVD were compared with those obtained by a classical method of least squares. They have been used a variety of software, including software written by author, also packages Matlab and Statistica. The main purpose of discussion is to solve sample tasks for better understanding and expand the use of decomposition SVD in geodetic issues.